Peephole: Predicting Network Performance Before Training

TL;DR

This paper introduces Peephole, a method that predicts neural network performance from architecture alone using layer encoding and LSTM, enabling reliable performance estimation before training.

Contribution

It presents a novel unified layer encoding approach combined with LSTM to predict network performance pre-training, addressing the challenge of large design space and training costs.

Findings

Achieved accurate performance predictions across multiple datasets.

Produced consistent architecture rankings.

Demonstrated the method's reliability and generalization.

Abstract

The quest for performant networks has been a significant force that drives the advancements of deep learning in recent years. While rewarding, improving network design has never been an easy journey. The large design space combined with the tremendous cost required for network training poses a major obstacle to this endeavor. In this work, we propose a new approach to this problem, namely, predicting the performance of a network before training, based on its architecture. Specifically, we develop a unified way to encode individual layers into vectors and bring them together to form an integrated description via LSTM. Taking advantage of the recurrent network's strong expressive power, this method can reliably predict the performances of various network architectures. Our empirical studies showed that it not only achieved accurate predictions but also produced consistent rankings across datasets -- a key desideratum in performance prediction.